Stem cell gene therapy remains an elusive therapeutic goal. Furthermore, without strategies to select for low levels of genetically altered stem cells expressing the gene of interest in vivo, it may be difficult if not impossible to impact hematopoiesis. The investigators have analyzed mutant forms of MGMT that provide resistance to the MGMT inhibitor, 06 benzylguanine (BG) and confer over 5 fold resistance to BG + BCNU in vitro and over 300 fold survival and hematopoietic reconstitution advantage in vivo following therapy with BG and BCNU. In the first Specific Aim, the investigators will evaluate human lentiviral vectors because of their enhanced ability to transduce human repopulating stem cells. Lentiviral backbone modifications will be made to increase transduction and expression, and mutant MGMT will be integrated to evaluate transduction into human CD34 and NOD/SCID repopulating human cells. In Aim 2, a dual gene vector will be evaluated containing the X linked Chronic Granulomatous Disease gene, gp91phox and MGMT. Transduction of murine stem cells from mice defective in gp9lphox followed by transplantation and non-myeloablative treatment with BG + BCNU will establish whether the MGMT gene can be used to select for a therapeutic gene in vivo. The third Aim is admittedly novel: the use of the MGMT gene to select for marrow-derived stem cells in nonhematopoietic organs such as liver, brain and lung. Since the preliminary studies by others indicate the plastic differentiation capacity of hematopoietic stem cells, we hypothesize that selective enrichment through a drug resistance strategy offers a potentially potent approach to tissue replacement. Establishing that a BG-resistant mutant MGMT gene can act as a therapeutic drug selection gene to protect stem cells and to enhance tissue engraftment in vivo has far reaching implications for stem cell gene therapy.